• State
  • Locked Locked
  • Replies 20 replies
  • Subscribers 65 subscribers
  • Views 2650 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

UCC28600 in non isolated feedback

saurabh kumar
Prodigy 160 points
Other Parts Discussed in Thread: UCC28600, TL431, UC3842

Hii,

If i want to use UCC28600 with non isolated feedback, what are the consideration to keep in mind while designing.

Also does UCC28600 trips when the voltage on feedback pin exceeds certain limit

  • 0
    TI__Guru* 86266 points
    Hello,

    I have not designed with this device in a non-isolated feedback converter but believe you should be able to do this with this device.

    The device was designed to operate with a TL431 opto isolator feedback scheme. There is a 20 K ohm internal resistor pulled up to an internal reference of 5V. You will require an external operational amplifier setup in a type 2 configuration. I reviewed the block diagram and there is a peak current limit threshold of 1.2V and there is a 1:2.5 buffer stage to the PWM comparator from the FEED back pin. When the FB pin is at 3V this should be maximum duty cycle. There also appears to be 400mV level shift circuit form the CS pin to the PWM comparator. So when the FB pin is at 1V this should represent 0V at the CS pin. You might want to find and operation amplifier whose common mode range is 0.5 to 4V. Also, clamping the amplifiers output from 0.25V to 4V is a good idea. When setting up the rails of the amplifier, you want to make sure the output cannot be pulled below ground.

    If anyone on the e2e has done a reference design with a non-isolated flyback converter. Could you forwarded the schematic to Saurabh to make his design process easier.

    Regards,
  • 0 in reply to Mike O'
    TI__Guru* 86266 points
    Hello,

    I reviewed the data sheet and figure 11 shows the controller operation based on the voltage at the feedback pin. You might find this helpful

    Regards,
  • 0 in reply to Mike O'
    TI__Guru* 86266 points
    Hello,

    You also might consider using a UC3842. It is a peak current mode controllers that can be used for flyback converters and has an on board operational amplifier.

    Regards,
  • 0 in reply to Mike O'
    Prodigy 160 points
    Actually i have built up a non-isolated feedback design, which is working fine at 230V AC & 160V, but The problem is that when the mains voltage is reduced to 160V AC , and switch on power supply the chip goes into restart mode. This problem is not being observed if the power supply is switched ON at 230V and mains input is reduced to 160V.
    I did observe the voltages at FB pin and drain current. At FB pin peak voltage level is 1.2V and on time of drain current is 10uS. I suppose is it this because transformer is getting saturated when PS is switched on at 160V or something else.

    Also if i add a fixed load on the output winding having feedback. then i dont observe this issue.
  • 0 in reply to saurabh kumar
    TI__Genius 15910 points

    If the on-time is 10us under these conditions then I'm guessing that the device is trying to switch at 40kHz under these conditions and that frequency is not adequate for the energy transfer required. 

    Can you please check the date code on your device (the date code would most likely start with a 5 and be something like 57M or 5AM.  It sounds like the part has not fully transitioned out of start up mode (all devices start at 40kHz during SS and then once FB takes control the part should regulate at the best switching frequency for that operating point.  The most recent devices have made the internal comparator to improve headroom and make the part more robust and avoid this condition...but first we need to verify what revision you have.  The solution could be as simple as getting the most recent part with its more robust operation.

  • 0 in reply to Lisa Dinwoodie
    Prodigy 160 points
    Hi Lisa,

    On the device there are two types of code.

    48M and AD35.
    Hope this would help you
  • 0 in reply to saurabh kumar
    TI__Genius 15910 points
    Yes, that is very helpful; the 48M is the date code and the AD35 is some kind of lot trace code.
    The important thing is the date code and yours tells me the part you have is from the previous revision (Rev A).
    A PCN (part change notice) was released in the later part of last year to announce that there is a Revision C released and it has been made more robust to prevent some issues, notably at the extreme temp ranges but also due to noise on the comparator rail, that would account for the issue you are seeing.
    Basically the bottom line is the part has not been changed in its functionality but is a hardier, more user friendly version and you should really get this Revision.
    Do you have a TI Field contact that you can get the more recent part from? You need to get devices that have the date code 5Ax, 5Bx, 5Cx, or anything that begins with a 6xx (the last digit, M or whatever, isn't as important as the first 2).
  • 0 in reply to Lisa Dinwoodie
    Prodigy 160 points
    Ok,

    So if i replace this with new part having date code as mentioned above, the problem would be solved. That would be nice. I will arrange some new parts and retest the same

    Kindly if you could elaborate more on the improvement done in updated parts, it would be helpful
  • 0 in reply to saurabh kumar
    TI__Genius 15910 points
    IF the problem is still there after you replace the old revision with the new revision then we will be talking once again (complete with waveforms and schematics...) but it has been my experience that your symptoms are exactly the reason why the part was re-spun and improved so it is the easiest first line of defense (but not a miracle if there are true application issues in your design).

    The main issue addressed was to correct a weak pull up node on the UVLO level shifter and internal power rail so that it would be more robust at all corners of operation, notably over the entire temperature range.
    Also fixed was the bias current generator circuit where the circuit was made less sensitive to noise by adding internal capacitance to make the current generator circuit more robust.

    Once you get the new parts (Rev C), please feel free to let me know how your circuit behaves so we can confirm that the issue is fully resolved.
    Best of luck!
  • 0 in reply to Lisa Dinwoodie
    Prodigy 160 points

    I have changed the part with new series, 5M series but the problem still exists.

    One more observation is when i power up the bias winding externally with 15V supply their is no problem in PS functionallity

  • 0 in reply to saurabh kumar
    Prodigy 160 points
    I am checking the switching frequency of my design under this condition, and it is coming to be 400Hz. and when i increase the load by addition 50mW, it increases to 2Khz.

    But the minimum switching frequency cant go below 32Khz. Strange!!!
  • 0 in reply to saurabh kumar
    TI__Genius 15910 points

    can you send the schematic and transformer specification? 

    the only time the controller can switch at less than 32kHz min (40kHz typ) is in burst packets at light load and even then it will have groupings of 40 kHz pulses.

    Is this issue occurring on multiple units?  Have you tried testing a different board?

  • 0 in reply to Lisa Dinwoodie
    TI__Genius 15910 points
    please send any scope plots showing VDD, CS, AUX winding (don't put a probe directly on OVP), FB, SS....in addition to the schematic and transformer spec.
  • 0 in reply to Lisa Dinwoodie
    TI__Genius 15910 points
    Hi Saurabh,
    Just checking to see if you have had a chance to get the scope shots and if you could send schematic and transformer info so we can dig into this.
  • 0 in reply to saurabh kumar
    Prodigy 160 points

    Yes, the root cause of the problem has been traced out.

    My primary inductance of transformer was on the higher side, 8mH due to which the ON time at lower input range was going in the order of 10uS.

    So i reduced the primary inductance to 4mH , which inturn eliminated the issue.

    Regards

    Saurabh

  • 0 in reply to saurabh kumar
    TI__Genius 15910 points
    Yeah!!! So glad to hear that!
  • 0 in reply to saurabh kumar
    Prodigy 160 points

    My primary to bias turns ratio: 4.7. Primary inductance: 5.6mH.2nd o/p  with feedback turns ratio:16 bias voltage 15v second o/p voltage is 3.7v. At 239v  I observe that fr 4 sec bias voltage is stable and then it drops to 6.5v/what could be the probable cause

  • 0 in reply to saurabh kumar
    TI__Genius 15910 points
    What is the switching frequency at the line and load conditions when the output falls? If it is around 40kHz then the inductor isn't sized right and needs to be smaller. It sounds like it is power limiting.
  • 0 in reply to Lisa Dinwoodie
    Prodigy 160 points
    • O/P falls at 125V AC input with o/P power of  0.5W. Switching frequency is 38KHz

    I cannot lower the primary inductance because i have to keep a fixed number of primary turns, around 200 with this turns to achieve a lower inductance lesser than 5.5mH i need to increase the air gap >0.3mm, for which no standarzed core is available.

    Can i have some means to tune the feedback circuit to increase the switching frequency. Because it would be difficult to lower the inductance value.

    Regards

    Saurabh

  • 0
    TI__Intellectual 1470 points
    Hi
    Please go through PMP4043 design.This reference design uses the UCC28600 quasi-resonant flyback controller to generate a 24V/0.1A output from a universal AC input source. This flyback converter is not isolated. Three transistors are used in place of an opto-coupler to reduce the overall BOM cost.
    just enter PMP4043 in the search option on www.ti.com.
    Regards,
    Harmeet